Maximum Likelihood Recursive Identification for the Multivariate Equation-Error Autoregressive Moving Average Systems Using the Data Filtering

Abstract

The maximum likelihood principle has wide applications in system identification. This paper studies the maximum likelihood identification problems of the multivariate equation-error systems with colored noise. The system is broken down into several subsystems based on the number of the outputs. The key is to transform the subsystem into a controlled autoregressive moving average model and a noise model. Based on the maximum likelihood principle and the data filtering technique, a filtering-based maximum likelihood recursive generalized extended least squares algorithm is presented for estimating the parameters of these two models. For comparison, a maximum likelihood recursive generalized extended least squares algorithm is presented. Finally, the simulation example results confirm the effectiveness of the two algorithms.